On February 11th, 2026, a coalition of Ethereum’s brightest minds—from the Foundation’s dedicated zkEVM and Protocol Support teams to client developers and the independent researchers at Ethproofs—convened for a landmark event: the first official L1-zkEVM workshop. This wasn’t just another technical meeting; it marked the formal kickoff of a plan to fundamentally rewire the heart of Ethereum. The goal is to replace the slow, costly, and redundant process of re-executing every transaction with a revolutionary alternative: verifying a single, mathematically irrefutable cryptographic proof. This shift, years in the making and now codified in a concrete 2026 roadmap, could finally deliver on the promise of a truly scalable and decentralized mainnet, but the path is fraught with unprecedented complexity and existential risks.
The End of Re-Execution: A New Era for Validation
For its entire history, Ethereum’s security has relied on a simple but brutally inefficient principle: every validator re-executes every single transaction in a block to confirm its validity. This “re-execution” model, while robust, is the primary bottleneck that has historically capped L1 throughput at a meager 15-30 transactions per second (TPS). The L1-zkEVM proposal aims to shatter this limit by integrating zero-knowledge proofs directly into the core protocol, a concept often referred to as “snarkifying the L1.” For a refresher on the current model, see our guide on How Ethereum Works.
Instead of thousands of nodes burning cycles on the same computations, a specialized actor called a “prover” will generate a zk-SNARK (Zero-Knowledge Succinct Non-Interactive Argument of Knowledge) for an entire block. This single, tiny proof mathematically guarantees the correctness of all transactions within it. Validators will then only need to verify the proof, a process that is exponentially faster and cheaper. This new workflow is the core of the 2026 initiative:
The Data: De-Risking Performance
For years, the primary objection to an L1-zkEVM was performance. Proving was considered too slow and expensive to be practical for mainnet blocks. However, 2025 saw a dramatic shift, as documented by the benchmarking platform Ethproofs. The ecosystem has now largely “de-risked” the performance question, with multiple zkVMs approaching or exceeding the North-star goals set by the Ethereum Foundation.
| Metric | North-Star Goal (2026) | Actual Progress (Dec 2025) | Improvement |
|---|---|---|---|
| P99 Proving Latency | ≤ 10 seconds | 15.9 seconds | From 16m 44s (62.5x) |
| Prover CAPEX | ≤ $100,000 | Line of sight for 10kW hardware | Feasible with consumer GPUs |
| Proof Size | ≤ 300 KiB | Varies by zkVM | Ongoing optimization |
| Average Proof Cost | N/A | $0.0376 | From $1.69 (45x) |
“2025 meaningfully de-risked zkEVM proving performance. We now have line of sight to a diverse set of n≈5 zkEVMs maintaining P99 <10 seconds on 10kW hardware. This shifts the center of gravity for 2026: from raw speed → toward quality, security, diversity, and resilience.”
Will Corcoran, Ethproofs [1]
The 2026 Roadmap: Six Tracks to a Snarkified Mainnet
The workshop, led by Kevaundray of the EF’s zkEVM team, detailed a concrete six-track roadmap for 2026. This is a complex, multi-stage effort involving tight coordination between Execution Layer (EL) and Consensus Layer (CL) client teams, zkVM developers, and security experts.
| Track | Description | Key Contributors |
|---|---|---|
| 1. Execution Witness & Guest Program | Standardizing the data (witness) EL clients produce for stateless validation and the program that consumes it. | Ignacio (EF), Peter (STEEL team) |
| 2. zkVM-Guest API Standardization | Creating a common interface so different zkVMs and guest programs are interchangeable, ensuring client diversity. | Marcin (EF) |
| 3. Consensus Layer Integration | Modifying CL clients (like Prysm, Lighthouse) to verify ZK proofs instead of re-executing blocks. | Francesco (EF), Manu (Prysm) |
| 4. Prover Infrastructure | Building the open-source infrastructure (like Ere, zkBoost) for generating and gossiping proofs across the network. | Han (EF), Stefan (PandaOps) |
| 5. Benchmarking & Metrics | Establishing rigorous metrics to track performance, inform gas repricing, and define hardware requirements. | Ignacio (EF), Fara (Ethproofs) |
| 6. Security & Formal Verification | Formal verification of all critical components (guest program, zkVMs, provers, verifiers) and continuous fuzzing. | Cody (EF), Alex H (EF), George (EF) |
The Bull vs. Bear Case: A High-Stakes Bet
The bullish case, championed by many within the Foundation, is that a successful L1-zkEVM integration would be the single largest scalability and decentralization upgrade in Ethereum’s history. By bringing the scaling benefits of Layer 2 solutions directly to L1, it could drastically lower gas fees and enable a new class of applications. Furthermore, it strengthens decentralization by lowering the hardware requirements to be a fully verifying node. A critical dependency is the upcoming “Glamsterdam” hard fork, which is expected to include Enshrined Proposer-Builder Separation (ePBS). Without ePBS, the time window for a prover to generate a proof is a mere 1-2 seconds; with ePBS, that expands to a much more manageable 6-9 seconds, making real-time proving feasible.
However, the bearish case is equally compelling and voiced by seasoned protocol observers. This is arguably the most complex technical challenge Ethereum has ever faced. A bug in the core prover/verifier logic could be catastrophic, potentially leading to an invalid state transition being accepted by the entire network—a failure mode far more severe than a smart contract hack. As one external auditor noted, “The formal verification track is not a ‘nice to have’; it’s the whole ballgame. And formally verifying something this complex is a multi-year, multi-million dollar effort with no guarantee of success.” Furthermore, there’s the economic challenge: how do you design a sustainable, decentralized incentive system for provers without creating new vectors for centralization or MEV extraction? Finally, some argue this is a case of “too little, too late.” With the L2 ecosystem already thriving and handling the majority of user transactions, is a multi-year effort to upgrade L1 a distraction from more pressing issues like improving the user experience of Account Abstraction across rollups?
Final Thoughts: Ethereum’s Identity Crisis
The L1-zkEVM workshop was more than a technical presentation; it was a statement of intent. The Ethereum Foundation is placing a massive bet that native, on-chain scaling is not just possible, but essential for the long-term security and decentralization of the network. The roadmap is ambitious, the technical hurdles are monumental, and the stakes could not be higher.
This initiative forces a difficult question about Ethereum’s future identity. Is it destined to be a secure, decentralized settlement layer for a vibrant ecosystem of L2s, or can it—and should it—also be a scalable execution layer in its own right? The 2026 roadmap is a direct attempt to answer that question with a resounding “both.” As the six tracks progress, the entire crypto world will be watching to see if Ethereum can pull off its most audacious upgrade yet, or if it’s chasing a dream that the rest of the ecosystem has already moved on from.
References
Will Corcoran, “Ethproofs 2025 Review & 2026 Roadmap,” HackMD
“L1-zkEVM breakout #01, Feb 11, 2026,” Fellowship of Ethereum Magicians
